The new MSP extension to the Max programming environmentprovides an easily comprehensible and versatile way to programrealtime DSP applications. Because of its full integration intoMax, MSP allows one to combine MIDI data and audio datareadily in any program, and to hear the results immediately. Thismakes it an excellent environment for experimenting with newDSP algorithms and for designing music performances with arealtime DSP component.This paper presents some algorithms for time-domain audioprocessing in MSP which are not commonly found in therepertoire of included effects for commercially available audioprocessors. These algorithms—which use the realtimesegmentation of captured audio—are computationallyinexpensive, yet are capable of producing a variety of interestingsonic effects. They include simulated time-compression andpitch-shifting of audio samples, segmentation of audio samplesfor use as “notes” in another rhythmic structure, and modulationto extreme rates of sample playback.
